I believe entrepreneurship is building your ship in the middle of a storm, so that you sail to the shore of success. After one such stormy day, when I thought will end the night on a high note by checking the Google Analytics dashboard for our product website. I was already very confident of the traffic numbers as the team was keeping track of the ad campaign dashboard and it was showing strong website click numbers. This entire process of checking the analytics dashboard was to make myself feel good before I go to sleep after a hard day. But what followed was never expected…
Alright to get perspective it’s important to go back a little to understand how important is this…
Building a brand through online advertising has always been under shadow of doubt. And most big brands till few years back even doubted the power of digital to get ROI otherwise. Hence, digital advertising was considered just a tick box media in the media plan. But, things have changed now, at least that is what I was told.
We ran a campaign for the launch our new product BECK Friends to get sign ups on www.beckfriends.com. We were positive that the only way to build the brand online and get incremental sign ups is using digital platforms. And when we say digital platforms then definitely the way to go is using Google & Facebook ad platforms, this covers 80% of the traffic on web and also it is not surprising that these two companies contribute to the 80% of the spends online in India! This is a worrying trend, but more on that maybe next time.
For now the problem at hand is something else. Something bigger maybe not for the big companies, but definitely for a small bootstrapped startup like us. We are self funded which means we are trying to build a brand get users to sign up & try our product with our own money. This means our budgets are low so every click count. And that’s exactly why we use digital advertising extensively, because NEWS!!! NEWS!!! Digital Advertising is where you can track every click.
Now with little experience we knew what to expect from social media advertising. Low cost clicks, but not necessarily converting into sign ups. So, we tried to divert traffic at low cost to the website and hook up the user with the value proposition we have with the product. But, we never expected what we got!
Cut to reality…
I check the Google Analytics dashboard for our website basis of a high performing campaign we are running on Facebook over the last ten days. To my shock every 100 Clicks charged on the ad campaign has only led to only 20 visits on the website.
The first feeling that it is a goof up at our end, so you start checking on the analytics tracking code on the website then move to checking the download speed of the website. All seems ok. We also add a second analytics code on the website to cross check, but it throws up similar numbers as the Google dashboard. So you try to research on the problem to figure out that there are many who are facing this issue and usually the discrepancy is 20-25%, which I feel is high too. But in our case it was 80%
Only thing left now was to shoot a mail to Facebook customer support to find out the reason. After which I get a prompt call from them where a very courteous customer support person says she understands and recommends that I should pause the campaign for now. This is followed by what seemed like a standard template email reply stating nothing conclusive, but suggested us to use unique tracking click thru URLs that can track the website traffic generated from the ad campaigns better. We do that, but still after couple of days of running the campaigns find the discrepancy in the numbers are the same.
Leaving us confused and staring at the two lines mentioned in the earlier email: “Kindly note that we aren’t able to provide any support or verification for any third-party tracking implementation or data collected. Our sincere apologies for the inconvenience.” In a way washing away from any responsibility of any variation that may occur in the tracking.
Come on guys you are Facebook, we look up to you when we build products. Advertising is from where you generate over 90% of your revenues and tracking the performance of the ads can’t be broken.
I have been the biggest endorser of digital advertising all these years. And as we all know that digital advertising is a complex subject as there are many factors that surround it. But one thing that sells that digital advertising majorly is because the performance can be tracked.
But, now this experience has put a big question in my head: How authentic are the numbers that we see on our campaign dashboard?
—
Deep Malhotra is the Co-Founder of www.beckfriends.com and Gemini Group. Deep has an entrepreneurial and startup experience in working & building companies like Google, MySpace.com & rediff.com.
Feel free to write with your thoughts and opinions on deep@imgemini.com or follow on Twitter @deepmalhotra or Linkedin: https://in.linkedin.com/in/deepmalhotra
Alok Rodinhood Kejriwal
Sorry to break your bubble, but that’s the way web conversions happen. Questions:
– what is your mobile vs. Online traffic?
– have you used many creatives and optimized on the best?
Your PROBLEM is the reason why Facebook came up with Instant Articles and Google came through with AMP! The “disappearance” of visitors are worst for ads but also for content
In case this helps – we never advertise but for our own “in-house” ads, we LOSE 20% clicks when they arrive at the landing page
Prateek Shah
There has always been a difference between facebook clicks and google analytics. Also, clicks on ad may go to the company page or be counted in the form of likes and shares, doesn’t necessarily lead to website.
Shikha Pandey
So I am sure the online guys would come to defend and demystify the way online/facebook ads work but what doesn’t show doesn’t exist. Where r the people going after clicking? No wonder when the startups raise money, most of the time it is diverted towards the marketing.. no bootstrapped startup is reaching anywhere with numbers like these
Shobhit Bakliwal
After clicking on your site’s link and waiting for more than 5 minutes to open your site, my patience was over. I refreshed and waited again, and still the loading icon keeps going round and round.
The problem seems to be with your site and not FB here. If I click on your ad and the page doesn’t open, why should I give it so much time? Google analytics registers a user after the page has loaded, which clearly isn’t happening because people are abandoning the site.
Fix that and you’ll get a huge chunk of your visitors back.
EDIT: Holy Fuck. The site finally opened (just got visible) in 9 minutes according to dev consoles. It is still loading resources. The biggest WTF moment to see that the background image you have added was added as a BASE 64 string, put in directly to your HTML. This increases the response time to first call to a huge number and then it is analyzed to start loading other files and resources.
Mayukh Basu
Hi Shobhit
Thanks for your response and for being interested enough to go through the code.
Firstly to address your issue that the website loads in 9 minutes,we are attaching screenshots to show the time our website takes to load on a desktop. You can see that the load time in this particular instance is 11 seconds. We have generally observed the value to be between 10-15 seconds. We would have to say that this is an issue with internet connectivity on your end.
Secondly about the image encoded as a BASE 64 string in the html file, that’s something we incorporated to reduce the number of HTTP calls made(and the implicit latency).
Deep Malhotra
Load time snapshot..
Sourabh Jain
Hi Deep
On Facebook dashboard, do the numbers for “clicks” and “link clicks” differ? Because “clicks” includes post likes, comments, shares etc. while “link clicks” means actions that take user away from facebook, like to the website you are promoting – just my 50 cents.
Bala Satish G
1. clicks and link clicks are different things. FB can only count the clicks on its interface which is different from website clicks.
2. Google analytics will only record the view once page load is complete, Check your mobile page load time. If user leaves intermittently before the js loads it will not be counted.
Atul Shah
True..Social media is too crowded to get the Startups noticed. A lot of startups I talked to are facing the same problem. They are realize that a lot of cash needs to be burnt to get through the clutter on Online Channels. A new convergent medium is required that has benefits of both…Online (targeted audience, RoI measurement tools and low cost) and Offline (Trust and Brand Recall).
Deep Malhotra
Hey Sourabh,
Though the dashboard doesn’t show a breakup no. for website & link clicks.
still if we would add those clicks to the analytics nos. they don’t even come close to the FB dashboard nos.
the differential of 80% in nos. is HUGE gap to fill..& we have no answer to it..
Deep Malhotra
Thanks Bala.
We agree! we have accounted for those clicks. Difference of 20-30% in dashboard is acceptable (though still very high according to me), but not 80% for sure..
Even for offline media it is said: “Half the money I spend on advertising is wasted; the trouble is I don’t know which half.” in this case FB ads wasted 80% of the spends..& I don’t know where it went..neither does FB
Deep Malhotra
Thanks Atul. Integration of Media will definitely give better ROI but problem here is of tracking by FB & I believe they earn their bread & butter thru online ads.. such discrepancy shouldn’t be there for their ad campaigns in the first place..
Gaurav Mishra
Are you tracking the links clicked? Use Bit.ly or any other similar tool and see if that makes any difference.
Yatin Mulay
How are you tagging your links in the campaign? Can you paste one of the URLs here?
Facebook is not consistent with the referrer data ( there are multiple referrers used by FB ). Referrer data can drop sometimes while being passed from a HTTPS site to a HTTP site, normally 20% of the time is common for any site that is typically a HTTP and not a HTTPS. However, 80% is a clear outlier case.
(For the traffic originating from the secure sites, normally the referrer is not passed to the destination site since there is no referrer present in the header, but sites like FB, Google, Twitter etc use a temporary redirect to http so as to preserve the referrer.)
Since your site is also on HTTPS the chances of referrer dropping are minimal ( HTTPS can talk just fine with another HTTPS site in terms of preserving the referrer data )
Your 80% attribution delta is most likely a tagging issue and NOT facebook issue or Google analytics issue. If your FB campaign URLs are not tagged with UTM parameters, analytics can not necessarily assign it a proper medium and will throw your data in the default data graveyards namely “DIRECT” or channel “Other”
If you tag your URLs consistently with UTM parameters, i’m 101% sure your problem would be resolved to at least 90-95%
In the meantime, the things you can do towards your data reconciliation efforts:
a) Check the Source/Medium under All Traffic . Filter for source = facebook and then see if you notice any unusual referrer as medium
b) Check the “Other” channel under the Channels section. Apply a secondary dimension of source/medium and see if any entries are present with source as facebook and medium either as ocpm or cpc or cpm. I’m guessing you’d find a lot of facebook/ocpm laying orphan here.
The default channel grouping in analytics takes care of only cpc and cpm, not the ocpm (Optimized CPM). If you wish you can change your default channel grouping under the Admin section of analytics or even create a customized one. The simplest solution would be properly tagging all your campaigns with UTM parameters ( consistency is the key ).
Use either Google URL builder Tool or better still Cardinal Path’s Campaign tagging tool ( it’s free ) to keep all your tagging centralized and organized neatly ( this is what i also use personally )
Hope this helps. Cheers!